nanoepiseg 1.0.0b3

Methylome segmentation algorithm using a changepoint detection HMM

NanoEpiSeg 1.0.0b3

NanoEpiSeg is a tool for de novo segmentation of a methylome from read-level methylation calls (such as Nanopolish).

NanoEpiSeg is currently in development. Please do not hesitate to report bugs or feature requests.

A detailed documentation is in the works, stay tuned!

Prerequisites

NanoEpiSeg assumes that your methylation calls are stored in MetH5 format.

Installation

Through pip:

pip install nanoepiseg

Through anaconda:

conda install -c snajder-r nanoepiseg

Usage

NanoEpiSeg is meant to be parallelizable, which is why you would typically call nanoepiseg on a cluster system in parallel. In order to best accomplish load-balancing, it is done per hdf5 chunk.

You can list the number of chunks per chromosome via:

nanoepiseg list_chunks --m5file INPUT_FILE.m5

To then perform segmentation for a certain chunk, you can run:

nanoepiseg segment_h5 --m5file INPUT_FILE.m5 \
    --out_tsv OUTPUT_FILE \ 
    --reader_workers NUM_READER_WORKERS \ 
    --workers NUM_SEGMENTATION_WORKERS \ 
    --chromosome CHROMOSOME \
    --chunks CHUNK1 [CHUNK2 ...]

There are further options available. Please check out the help to discover them.